Table 2.
Preliminary validation of the H5 EB-ELISA
| Validation criteria and the samples used | Samples [N] | Assays [n] | Results | ||
|---|---|---|---|---|---|
| Analytical specificity (Asp) | True negative | False positive | Asp (%) | ||
| TN | FP | TN/(TN + FP) | |||
| Reference antisera against non-H5 subtype LPAIVs | 22 | 43 | 43 | 0 | 100 |
| Diagnostic specificity (Dsp) | True negative | False positive | Dsp [%] | ||
| TN | FP | TN/(TN + FP) | |||
| Various-origin sera negative against H5 HA | 209 | 209 | 204 | 5a | 97.6 |
| Diagnostic sensitivity (Dse) 1 | True positive | False negative | Dse 1 (%) | ||
| TP | FN | TP/(TP + FN) | |||
| Reference antisera against H5-subtype LPAIVs | 9 | 99 | 97 | 2b | 98.0 |
| Diagnostic sensitivity (Dse) 2 | True positive | False negative | Dse 2 (%) | ||
| TP | FN | TP/(TP + FN) | |||
| Experimental antisera against HA from H5N1 HPAIV | 115 | 115 | 114 | 1c | 99.1 |
| Repeatability of assays | Mean [OD450] | SDd | RSDe (%) | ||
| G-7-27-18 mAb control | 1 | 13 | 1.566 | 0.111 | 7.1 |
| Normal chicken serum (negative control) | 1 | 13 | 1.460 | 0.114 | 7.8 |
| Anti-H5N2 LPAIV antiserum (weak positive control) | 1 | 12 | 0.944 | 0.083 | 8.8 |
| Anti-H5N3 LPAIV antiserum (strong positive control) | 1 | 12 | 0.439 | 0.044 | 10.0 |
Serum samples (Table 1) were classified as anti-H5 HA positive or negative based on the HI assay results. HI titers for the reference antisera were provided in the product certificates. Normal chicken serum and sera from commercial chickens immunized and nonimmunized with rH5-E. coli were analyzed in the HI assay according to the protocol included in the Methods section. Data for all experimental antisera and 96 out of 210 sera negative against H5 HA were adapted from our previous paper [10]. HI titers for the reference and experimental antisera against H5-subtype AIVs are provided in Table 1 and Additional file 1: Tables S2 and S4. In this study, serum HI titers equal to or greater than 1:8 were considered positive. The H5 EB-ELISA was performed, and inhibition percentages were calculated as described in the Methods section. Samples showing inhibition above the cutoff value of 38.5% were considered positive against HA of H5-subtype influenza viruses in the EB-ELISA. Analytical and diagnostic specificities were calculated by counting the samples determined in the EB-ELISA as true negatives among the anti-H5 HA negative sera from chickens immunized with LPAIVs of the H1–H4 and H6–H16 subtypes and the nonimmunized chickens, respectively. Diagnostic sensitivities 1 and 2 were calculated by counting the samples determined in the EB-ELISA as true positives among the anti-H5 HA positive reference and experimental antisera, respectively. The repeatability of the EB-ELISA was evaluated by performing the indicated number of assays for the listed control samples
aThe samples yielding from 39.7 to 43.0% inhibition
bTwo out of twelve assays of the anti-H5N2 LPAIV antiserum, batch #3, with an HI titer of 1:512
cThe sample with an HI titer of 1:64
dStandard deviation
eRelative standard deviation